Re: That old jpg v. RAW argument again...

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On Wed, 2004-09-15 at 12:53, Howard wrote:

> 
> Three points...
> 1	Can anyone actually see the difference in the final image between TIFF 
> and JPG at maximum quality (and not repeatedly saved and recompressed)??

I could probably see it, of course those artifacts are more visible in
some images then are in others, and in high quality settings are in fact
almost invisible, it would take a very careful examination to find it. 

But there could be a need to do some crop or color correction to the
image and this would imply in a recompression if you start with a
pristine raw copy this will have less impact. 

> 2	Can you actually see the difference between an 16 bit TIFF and an 8 
> bit TIFF??

16 bits of colors can be very useful to do some manipulation without
reducing the total number of colors in the end result. I never actually
used a 16bits image, but I can imagine a 8bit resultant image from a
color levels operation that would have a color histogram very similar to
an untouched image. 

Also with subpixel interpolation one can enhance the resolution of an
image while reducing the color resolution from 16bits to 8bits per
channel, with better results then if you simply enlarge it. 

> 3	And in any case, if the same image were shot simultaneously on Canon 
> 20D, Nikon D100, Pentax *st, Sigma 10D, Kodak etc., in RAW format and 
> converted to TIFF without any manipulation, I'll bet that these would 
> almost definitely show distinct differences when printed....
>
> Just as using different films gave differing results under as identical 
> a set of conditions as possible!
> 

The print operation is very lossy by nature. The screen and most
printers have a very different gamut of colors, and resolution. To
obtain good prints most printers do lots of manipulation in the image.
Even on photo paper, witch I am not very used to print on, I would guess
that similar pictures would result in very similar prints. Mainly
because there are many man-decisions in the print time (different
expositions). 

> So what DO we all mean by the "best quality"??

What do I want in a best quality? That's hard, low noise in higher iso
pictures is one thing. Good color reproduction. Raw 16bits images would
also make my day, but I did heard that the transfer speeds to the memory
chip can make shooting with those options prohibitively slow. 


[Index of Archives] [Share Photos] [Epson Inkjet] [Scanner List] [Gimp Users] [Gimp for Windows]

  Powered by Linux